Adapt-As-You-Walk Through the Clouds: Training-Free Online Test-Time Adaptation of 3D Vision-Language Foundation Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Tamjidi, Mehran, Dastmalchi, Hamidreza, Alimoradijazi, Mohammadreza, Cheraghian, Ali, An, Aijun, Saberi, Morteza
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908668050538496
author Tamjidi, Mehran
Dastmalchi, Hamidreza
Alimoradijazi, Mohammadreza
Cheraghian, Ali
An, Aijun
Saberi, Morteza
author_facet Tamjidi, Mehran
Dastmalchi, Hamidreza
Alimoradijazi, Mohammadreza
Cheraghian, Ali
An, Aijun
Saberi, Morteza
contents 3D Vision-Language Foundation Models (VLFMs) have shown strong generalization and zero-shot recognition capabilities in open-world point cloud processing tasks. However, these models often underperform in practical scenarios where data are noisy, incomplete, or drawn from a different distribution than the training data. To address this, we propose Uni-Adapter, a novel training-free online test-time adaptation (TTA) strategy for 3D VLFMs based on dynamic prototype learning. We define a 3D cache to store class-specific cluster centers as prototypes, which are continuously updated to capture intra-class variability in heterogeneous data distributions. These dynamic prototypes serve as anchors for cache-based logit computation via similarity scoring. Simultaneously, a graph-based label smoothing module captures inter-prototype similarities to enforce label consistency among similar prototypes. Finally, we unify predictions from the original 3D VLFM and the refined 3D cache using entropy-weighted aggregation for reliable adaptation. Without retraining, Uni-Adapter effectively mitigates distribution shifts, achieving state-of-the-art performance on diverse 3D benchmarks over different 3D VLFMs, improving ModelNet-40C by 10.55%, ScanObjectNN-C by 8.26%, and ShapeNet-C by 4.49% over the source 3D VLFMs. Project page: https://mehran-tam.github.io/Uni-Adapter
format Preprint
id arxiv_https___arxiv_org_abs_2511_15311
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Adapt-As-You-Walk Through the Clouds: Training-Free Online Test-Time Adaptation of 3D Vision-Language Foundation Models
Tamjidi, Mehran
Dastmalchi, Hamidreza
Alimoradijazi, Mohammadreza
Cheraghian, Ali
An, Aijun
Saberi, Morteza
Computer Vision and Pattern Recognition
3D Vision-Language Foundation Models (VLFMs) have shown strong generalization and zero-shot recognition capabilities in open-world point cloud processing tasks. However, these models often underperform in practical scenarios where data are noisy, incomplete, or drawn from a different distribution than the training data. To address this, we propose Uni-Adapter, a novel training-free online test-time adaptation (TTA) strategy for 3D VLFMs based on dynamic prototype learning. We define a 3D cache to store class-specific cluster centers as prototypes, which are continuously updated to capture intra-class variability in heterogeneous data distributions. These dynamic prototypes serve as anchors for cache-based logit computation via similarity scoring. Simultaneously, a graph-based label smoothing module captures inter-prototype similarities to enforce label consistency among similar prototypes. Finally, we unify predictions from the original 3D VLFM and the refined 3D cache using entropy-weighted aggregation for reliable adaptation. Without retraining, Uni-Adapter effectively mitigates distribution shifts, achieving state-of-the-art performance on diverse 3D benchmarks over different 3D VLFMs, improving ModelNet-40C by 10.55%, ScanObjectNN-C by 8.26%, and ShapeNet-C by 4.49% over the source 3D VLFMs. Project page: https://mehran-tam.github.io/Uni-Adapter
title Adapt-As-You-Walk Through the Clouds: Training-Free Online Test-Time Adaptation of 3D Vision-Language Foundation Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.15311